Background: Stroke is the primary cause of adult long-term disability and its incidence is predicted to rise with our aging population. Residual arm deficits persist in 55% or more of stroke survivors six months post onset. The upper extremity (UE) is an important target for rehabilitation as it impacts independence in activities of daily living (ADL) and quality of life. ur investigations utilizing robot-assisted therapy for intense repetition and task specificity have resulted in positive effects on reducing UE impairment in Veterans with moderate to severe deficits ? six months post stroke. An expansion of our work suggests that intensive robot- assisted reaching combined with a brief period of task specific (Transition to Task Training (TTT)) leads to greater gains than robot training alone. Current robotic devices are large and expensive to deploy in large numbers. An alternative shown to have positive effect in this population is Bilateral Arm Training with Rhythmic Auditory Cueing (BATRAC). This is a low- cost mechanical device based on similar motor learning features. Unlike the robot, it does not have an assist mode and requires engagement of the less impaired arm to work in concert with the affected arm. This bilateral practice may enhance inter-hemispheric coupling of the damaged motor cortex to promote spatial and temporal interactions of the arms not seen in other forms of repetitive training. Objective: We propose to adapt the lab-based BATRAC training into a home-based model with a telerehabilitation platform. We will add TTT since it appears to be quite powerful in promoting paretic hand use based on the Stroke Impact Scale Hand Section. Our primary objective is to compare the effects of 6 weeks of UE neuro-rehabilitation using Home-based BATRAC telerehabilitation with TTT on arm function in patients with chronic stroke to the usual and customary care provided at this time point. Additional objectives are to determine whether Home-based BATRAC telerehabilitation + TTT will be at least as effective as lab-based and whether patient-centered telerehabilitation has an effect on behavioral-psychosocial outcomes. Methods: Participants with UE deficits ? 6 months duration due to stroke will be stratified by impairment and randomly assigned to immediate therapy or delayed-entry therapy. The immediate therapy group will be randomly assigned to one of two study arms: (1) lab-based BATRAC+TTT or (2) a home-based BATRAC telerehabilitation for repetitive, dose-matched reaching+TTT. Similarly, after serving as a usual care control for 6 weeks, the delayed-entry group will be randomized to either study arm 1 or 2. When the delayed entry individuals have completed their interventions, the two groups (1 and 2) will be compared (after 6 week training and after 8 week retention) to the immediate therapy study arms. The primary outcome measure is the WMFT conducted by the blinded assessment team at baseline, week 6, and week 14. All assessments for each group will occur in the clinic setting. Results: Expanding the continuum of stroke care with a home telerehabilitation format has the potential to increase opportunity for intensive UE practice and translation of gains to real world use. When we have completed these aims we expect the results to provide an objective basis for expansion of UE training in neurological populations such as traumatic brain injury, multiple sclerosis as well as design a program that can be disseminated more broadly through the MyHealtheVet web system.